November 15, 2006;
ADMP2 is essential for primitive blood and heart development in Xenopus.
We describe here the cloning of a new member of the TGF-beta
family with similarity to the anti-dorsalizing morphogenetic proteins (ADMPs). This new gene, ADMP2
, is expressed in a broad band of mesendoderm
cells that appear to include the progenitors of the endoderm
and the ventral mesoderm
. Antisense morpholino oligonucleotide knockdown of ADMP2
results in near-complete disruption of primitive blood
development, while the development of other mesoderm
derivatives, including pronephros
plate is not disrupted. Moreover, the development of the primitive blood
knockdown embryos cannot be rescued by BMP. These results suggests that ADMP2
plays an early role in specifying presumptive ventral mesoderm
in the leading edge mesoderm
, and that ADMP2
activity may be necessary to respond to BMP signaling in the context of ventral mesoderm
[+] show captions
Fig. 1. The predicted ADMP2 protein. (A) Alignment of mature domain of Xenopus laevis ADMP2 (XlADMP2) with other members of the TGF-β family. (B) The schematic representation of the relationship between XlADMP2 and members of the ADMP and BMP families. The number at each branch point represents the relative incidence of that particular relationship during bootstrap resampling using 1000 replicates. The scale bar at the left bottom indicates an evolutionary distance of 0.1 amino acid substitution per position.
Fig. 2. Expression of ADMP2. (A) Expression of ADMP2 by Northern blotting of RNA from staged Xenopus embryos. Expression was first seen at stage 9 and persisted through the early neurula stage. Ethidium staining for 28S and 18S RNAs serves as a loading control. (B) In situ hybridization for ADMP2 in stage 10 embryo bisected sagittally. Arrows indicate general region of specific hybridization. (C and D) Two halves of bisected stage 10.5 embryo hybridized with either ADMP2 (C), or brachyury (D) (for the benefit of comparison, one image is inverted). (E and F) Sagittal (E) and para-sagittal (F) sections of stage 11 embryos hybridized for ADMP2. (G) Vegetal view of schematic stage 11 embryo indicating the approximate lines of bisection of embryos in panels E and F. (H and I) Stage 12 embryos hybridized for ADMP2. Dorsal is up and ventral is down for both embryos. Panel H shows sagittal section, while panel I shows transverse section. (J) Schematic view of stage 12 embryo showing the approximate plans of view in panels H and I. S.O.: approximate position of the Spemann organizer. bc: position of the blastocel.
Fig. 3. Regulation of ADMP2 expression in animal caps and whole embryos. Northern blot of RNA from stage 10.5 animal caps injected with 1 ng of Xnr2 with or without 400 pg of XFD RNAs. The figure shows the results of hybridizations for ADMP2, ADMP and EF1α (loading control). Xnr2 RNA alone induces a low level of ADMP2 expression in animal caps, while co-injection of XFD RNA causes a much higher level of ADMP2 expression. ADMP expression, on the other hand, was inhibited by XFD.
Fig. 4. Overexpression of ADMP2. (A) Northern blot of RNA from stage 32 sibling embryos of the ones shown in panels C–F and H–J. The figure shows the results of hybridizations for globin, muscle actin (m.actin) and EF1α (loading control). The expression of globin was increased by overexpression of ADMP2, ADMP or BMP7 RNA (lanes 2, 4 and 6 compared to lane 1), while that of m.actin was virtually eliminated by the same treatment (lane 2, 4 and 6 compared to lane 1). The level of m.actin expression in the BMP7-injected embryos was rescued to normal levels by co-injection of tBR RNA (lane 7 compared to lane 1). However, in the ADMP2 or ADMP-injected embryos, co-injection of tBR RNA failed to rescue m.actin expression (lanes 3 and 5 compared to lane 1). (B) The phenotype and globin expression by in situ hybridization in ADMP2-overexpressed embryos. The expression of a blood marker, globin, is normally detectable in the ventral-most region of the tailbud stage embryo at stage 32 (white arrowheads in top embryo). Overexpression of ADMP2 caused a typical DAI 0 phenotype (two bottom embryos). The expression of globin was restricted in the leading edge mesoderm of these embryos (white arrowheads) at the equivalent of stage 32. (C–J) Rescue experiments by co-injection of tBR RNA. Overexpression of either ADMP2, ADMP or BMP7 RNA caused a similar DAI 0 phenotype (C–E compared to B). However, only the phenotype caused by overexpression of BMP7 was rescued by co-injection of tBR RNA (G–I), suggesting that ADMP2 and ADMP do not bind the BMP receptor. (G) The phenotype of tBR RNA-injected embryos. All the embryos shown are at stage 32.
Fig. 5. (A) Inhibition of translation by an ADMP2 antisense morpholino oligonucleotide. Western blot of proteins extracted from stage 14 embryos that have been injected separately with 100 pg of RNA encoding an ADMP2/GFP fusion protein and either 10 ng of a control (c-MO) or antisense ADMP2 morpholino oligonucleotide (A-MO). Shown is the result of immunoblotting using an anti-GFP antibody and an anti-alpha-tubulin antibody (loading control). (B–G) Loss-of-function analysis of ADMP2. Control embryos (B, C, D), and embryos (E, F, G) injected with an antisense ADMP2 morpholino oligonucleotide (A-MO) probed for expression of a somite marker, XmyoD (B, E), a pronephros marker, Xlim1 (C, F) or a lateral plate marker, Xnr1 (D, G). (B, E) Stage 14 embryos. Anterior is to the left. Dorsal views. (C, D, F, G) Stage 24 embryos. Lateral views. Anterior to the left.
Fig. 6. Loss of ADMP2 and heart formation. (A and B) Injection of 10 ng of an antisense ADMP2 morpholino oligonucleotide into each of the two anterior blastomeres of a 4-cell stage embryo inhibited the expression of the heart marker Nkx2.5 (c-MO: control morpholino; A-MO: antisense ADMP2 morpholino). (C) This effect was reversed by co-injection of 5 pg of ADMP2 δ-5′ UTR RNA (total 10 pg) (black arrowheads). (D) Transverse sections of c-MO-injected embryos showed that Nkx2.5 was expressed in a tube of myocardium (white arrowheads) as well as the pericardial roof (black arrowheads in panel D). (E) A-MO-injected embryos, on the other hand, showed only faint expression of Nkx2.5 (white arrowhead in panel E), and the heart tube was not formed. (F) Rescue of the A-MO phenotype, as in panel C. (G, H) Anterior injection of 10 ng A-MO also reduced the expression of TnIC, another heart marker (black arrowhead in panel K compared to that in panel J) (82%; Table 1).
Fig. 7. ADMP2 loss and blood formation. (A and B) Posterior injection of 10 ng of antisense ADMP2 morpholino oligonucleotide (A-MO) into each of the two posterior blastomeres at the four cell stage inhibited the expression of globin in the posterior ventral blood islands (c-MO: control morpholino). (C) Co-injection of 10 pg of ADMP2 δ-5′ UTR RNA (total 20 pg) with A-MO restored globin expression (white arrowhead). (D–F) Anterior injection of 5 and 10 ng/blastomere of A-MO. While the 5 ng dose inhibited globin expression when injected posteriorly (E), no reduction in globin was observed at this dose anteriorly. At 10 ng (F) a strong reduction in globin expression anteriorly was observed. (G and H) The expression of SCL, an early blood marker, was also reduced by injection of 10 ng of A-MO posteriorly when examined at stage 24. (I and J) Transverse sections at the level of the posterior blood islands show that the expression of globin was decreased to a thin layer in the A-MO-injected embryos (black arrowhead in panel J compared to that in panel I). (K) In embryos co-injected with A-MO and ADMP2 δ-5′ UTR RNA, the globin-expression extended deep into the mesodermal layer. (L) Enlarged view of ventral mid-line showing globin expression in either c-MO (left) or A-MO-injected embryos. (M, N) Coinjection of 10 ng A-MO into each blastomere at four cell stage reduced globin expression in response to BMP7 RNA injection (1 ng per embryo). Note that both control (M) and A-MO-injected embryos (N) have the same DAI 0 phenotype.
admp2 (antidorsalizing morphogenetic protein 2) gene expression in bissected Xenopus laevis embryo, assayed via in situ hybridization, NF stage 10, horizontal view, dorsal right